Document 6RvwMk3Kmb1e0Q7BExyepbjzg

; i HP COPIED \ ctcrc. 53nq-54o] \ aMfli PyM/M'H'' n > i/'irwji Ctfitiricu iy(M^ h^mc-i' JAN 0 6 1992 The Dow Chemical Company 2030 Dow Center January 2, 1992 r7/o/ iZ/t/VJL. &MfUDE Ms. Kathy Rosicca Chemical Manufacturers Association 2501 M Street, Northwest Washington, D.C. 20037 "PATHOLOC?Y'ftEBpRT ON THE BRAINS FROM MICE, HAMSTERS ANt) RATS EXPOSED TO - VINYL CHLORIDE^ IBT STUDY 663-03222 - Submitted to Chemical Manufacturers AssociatiefrotTMay 14, 1980 by experimental Pathology Laboratories, Inc. 1 Dear Ms. Rosicca: Dow has entered into a TSCA 8(e) (Compliance Audit Program) agreement with the EPA. This is a one time voluntary audit program to locate and report any outstanding studies or other information reportable under this section. We currently have the above-described study in our file and feel it is reportable under section 8(e) of TSCA. We are forwarding a copy to the EPA and wanted to let you know of this submission. If you have any questions, please do not hesitate to contact me. Very truly yours, Legal Department (517) 636-4879 cc: R. L. Hagerman, 1803 Building J. E. LeBeau, 2030 Dow Center P. A. Wright, 2030 Dow Center wss to _/UL! ____ CORPORATE SAF^f JAN 30 1992 &N0 health deft. GENC 005379 Vinyl Chloride STUDY UNKS LIVER, BRAIN CANCER TO WORKING MEN EXPOSED TO CHEMICAL An industrywide study of workers exposed to vinyl chlo ride found they were significantly more likely to die from cancer of the liver and biliary tract, cancer of the brain and central nervous system, and emphysema and chronic ob structive pulmonary disease. The study, published in the September American Jour nal of Industrial Medicine (Vol. 20, No. 3), was based on data on 10,173 men who had worked for at least one year in jobs involving exposure to vinyl chloride before January 1973. The workers were employed at 37 U.S. plants belong ing to 17 companies. The researchers--led by Otto Wong of Applied n;th Sciences of San Mateo, Calif.-analysed length of exposure, latency, age at first exposure, calendar year of first exp<v sure, and type of products to which workers were exposed. There were 1,536 deaths identified in the study. Mortality from all types of cancer was slightly, but not significantly, higher than the average U.S. experience. But. there were 37 deaths from cancer of the liver and biliary tract--which the researchers referred to simply as liver--compared with only 5.3 expected. There were 23 deaths for cancer of the brain and other central nervous system, compared with 12.8 expected. No other cancer types showed any statistically significant ex cess, the study said. Two Plants Approximately half the cases of brain cancer came from taro plants that were manufacturers of only polyvinyl chloride--as opposed to producing vinyl chloride monomer, both PVC and VCM, or homopolymer and copolymers. The researchers said these two plants produced a "disproportion- /ff/ . CURRENT REPORT ately large number of deaths" from cancer of the brain and nervous system when compared to other plants in the study. Study results also indicated that for the entire cohort, 41 deaths were due to emphysema or other chronic obstructive pulmonary disease, compared with the expected 22.8. The excess came mainly from those with less than 10 years of exposure: in this group there were 25 deaths from emphyse ma and related illnesses. The researchers noted that their findings of a statistically v significant excess in emphysema mortality have not been reported previously by other investigators. No Smoking Data The researchers said because the study did not contain data on smoking history, they were unable to assess its role in increased mortality from emphysema. "However, it must be pointed out that there was no excess in lung cancer, which would imply that the cohort did not have a higher percentage of smokers than the general population," they said. The major limitation of the study was lack of exposure information, the researchers said. Because of this, "analyses by detailed exposure were not feasible," they said. Also, the researchers were unable to get vital status follow-up information on a small percentage of the cohort And finally, because it was a mortality study, the research lacked in-depth clinical information, the researchers said. In particular, histologic information on the cell type of the liver cancers and brain cancers "would have been useful in interpreting the epidemiologic findings," they said. GENC 005380 ZI\ 'IHFV^CA,. VIANU^CTijRERS ASSOCIATION November 20, 1991 Dear Vinyl Chloride Research Coordinators: Otto Wong et. al. recently published the enclosed article entitled, "An Industry-Wide Epidemiologic Study of Vinyl Chloride Workers, 1942-1982." The publication was based on studies that the CMA Vinyl Chloride Panel had sponsored in the past. Dr. Wong did not send a draft manuscript to CMA for review and comments prior to the publication. Drs. Bob Hinderer and J.R. Drumwright have expressed concerns regarding the interpretation of our data By Dr. Wong et. al. The enclosed letters from Drs. Hinderer and Drumwright describe their concerns. I would like to schedule a conference call during the week of December 1, 1991, to discuss the enclosed publication and the concerns expressed by Drs. Hinderer and Drumwright, and to decide on a course of action. Please complete the enclosed survey form concerning your availability and return it to me as soon as possible. You may fax your availability survey to me at (202) 887-1237. I have asked Dr. Bill Gaffey, who is very familiar with the studies we sponsored, to be available for the conference call. We may want to consider hiring him to prepare our response to Dr. Wong's publication. Sincerely, -- Hasmukh C. Shah Manager, Vinyl Chloride Panel cc: J.R. Drumwright William Gaffey James Barter Eugene Skiest A> A SP _Q. CORPORATE SAFETY NOV 22 1991 and HEAJ.TH 0PT. 2501 M Street. NW, Washington, DC 20037 202-887-1100 Panafax 202-887-1237 Telex 89617 (CMA WSH) GENC 005331 CHEMICAL MANFUCTURERS ASSOCIATION Vinyl Chloride Research Coordinators November 20, 1991 Availability Survey Please circle the day(s) and the time(s) that you are available for a conference call. I am available for a conference call on: a) Monday, December 2, 1991 (Morning/Afternoon) b) Tuesday, December 3, 1991 (Morning/Afternoon) c) Wednesday, December 4, 1991 (Morning/Afternoon) d) Thursday, December 5, 1991 (Morning/Afternoon) e) Friday, December 6, 1991 (Morning/Afternoon) NAME: _____________________________ COMPANY: ADDRESS: __ _________________ _ PHONE: FAX: PLEASE RETURN VIA FAX TO: Hasmukh Shah CMA FAX: (202) 887-1237 GENC 00538 OCT 8:59 FPQW CLE(J-HtTH-SC I -|_ I B 'LiGE , iwHcw Journal of Industrial MsdWos 2Wl7**J34 (1991) -S3. whins ud to. i pulmonary an Milbsafc Q rr Hasp Dit iowa Stats un lad Hyj ammonia m An Industry-Wide Epidemiologic Study Of Vinyl Chloride Workers, 1942-1982 Otto Wong, $cD, M. Donald. Whorton, wo, Donna Pollan, mo, and David Ragland, pho The cohort consisted at 10,173 toco who had worked for ti lean one year is job* ' involving exposures vinyl chloride prior to l January 1973. These men wen employed ( *V at ?7 plant* in tftt'TJ.y belonging to 17 companies. Observation of the mortality + experience of cohort WasUpdared from 31 December 1972 to 31 December 1982 (the study now covering 1942-1982). A total of 1.336 cotton members wen identified as having died. The observed mortaJUy, by cause, was compared with the expected baled on U.S. mortality rates, standardized for age, race, and calendar rime. Analyses by length of exposure, latency, age at first exposure, calendar year of first exposure, sad type of products were performed. The study confirmed that the vinyl chloride woritats experience s significant mortality excesses in angiosarcoma (13 deaths), cancer at the liver and biliary tract (SMR * 641), and cancer of the brain and other central nervous system (SMR - 180). In addition, the study also found s significant mortality extra* in emphysema/ehronic obstructive pulmonary disease (COPD) (SMR 179). On the other hand, the study did not find any excess in either respiratory cancer or lymphatic and Hematopoietic cancer. This study also found in increase in biliary net cancers, independent from liver cancer. Key words: angiosarcoma. Brer oncer, bran, cancer, chronic utetructlve pofammary dfasaaa BACKGROUND In 1974, three cases of angiosarcoma of the liver were repotted among workers employed at a vinyl chloride (VC) polymerization plant in the U.S. [Creech and Johnson, 1974]. About the same time, similar cancerous hepatic effects were reported in Laboratory animals exposed to vinyl chloride [Maltoni et al., 19811. Since then, a number of epidemiologic studies have confirmed die association between angiosar coma of the liver and occupational exposure to vinyl chloride [Tabershaw and Gaffey, 1974; Cooper. 1981; Fox and Collier, 1977; Waxweiler et ai., 1976; Jones et al., 1988]. In addition to angiosarcoma of the liver a number of other cancer sites have also been implicated, including the overall digestive system [TMriault and Allard, 1981; Chiazze et al., 1977; Smuievich et al., 1988], the respiratory system [Monson Applied Health Saeoce,. Sen Mjttoo. CA (O.W.). ENSR Health Science*. Alameda, CA (M.D.W., D.EJ.. D.A.). Address rcprint mqunta to Dr. Otto Wong, Applied Health Science*. 1S1 Second Arenue. Suits 628. P.O. Box 2078, Sin Mateo, CA 9*01. CORPORATE SAFETY Acceptad for publication Decamber 31, 1990. 1991 Wtky-Ltss, lee. NOV 2Z 1S9! AND HEALTH DEPT. GENC 005383 OCT g: 01 91 FsonClE^-^j^Cl-LlB rnJt , fcl <J" 1988], et al., iweiler iignam viewed -posure out the jj their sr were cancer power tad for Chemidy of i were scatae OW.up t, that narily entral really *fa v*. Epidemiologic Study at Vinyl Chloride Worker* 319 additional years of observation). This report summarizes the conduct, statistical it* suits, and epidemiologic interpretation of the updated CMA study MATERIALS AND METHOD Original Study The cohort consisted of 10,173 workers from 37 plants belonging to 17 com panies. (Since the original study has been completed, the ownership of several plants has changed.) There were 11 plants with 1,214 workers which produced only vinyl chloride monomer (VCM), 18 plant with 6,84$ workers which produced only poly vinyl chloride (PVC), 3 plants with 935 workers which produced both, and 3 plana with 1.176 workers which produced homopolymers and copolymers. Twenty-two of the 37 plants ate located in the South, with eight in the Northeast, six is the North Central, and one in the West. To be eligible for the cohort, male employee* at the 37 participating plants must satisfy the following criteria: exposure to VC for at least 1 year prior to 12/31/1972; and employed during or after 1942. In the original study, the designation of jobs involving exposure to VC was made by staff members at individual plana or by corporate industrial hygienists. In approximately two-third* of tbe study population, employment records of all employ ees at the plana were microfilmed by the original investigation team, and exposed individuals were identified through these records [Cooper, 1981]. In the remaining portion of the population, deoiled employment information on exposed individuals was provided by plant personnel on special forms. A small portion of these individ uals were identified in tbe study only by a study number, and no personal identifiers were available to the investigators. The period of time in which tbe employees of a plant were included in tbe study depended on the date it began making or using VC and also on the earliest date when personnel records were complete for all employees, whichever was later. This was done to eliminate periods when there was differential record retention of workers terminated, deceased, or retired, since such differential record retention might result in a potentially biased cohort. The end of observation in the original study was 31 December 1972. In an attempt to characterize historical exposures, individual jobs and job lo cations indicated by company personnel exposed to VC were graded by company personnel in terms of probable exposure on a scale of "high." "medium," end "low. '* These terms were used to describe relative exposures within each plant The original investigators acknowledged in their report of the study that this approach failed to reconcile differences among plant exposures [Tabershaw and Gaffey, 1974], In some plants it was likely that a "high" grading might have meant average expo sures in excess of 200 ppm, while in others, a "high" might have indicated average exposures of about 50 ppm. Actual levels were not known. Study Update The vital status of this cohort as determined by the original study investigators included 8.970 (92.7%) alive. 707 (7.3%) dead, and 496 (4.9%) unknown. These GENC 005385 m ormed ive or . five umber der to work*P*gc t. The to the mfor- latioc. iwn dr on for Death Social lies of ending vision he 7th x 7tb ia the tency, xed in indard iginal igmai : race : with in our on the jh the plants : hired proup. i. less years gth of # FROM CLEO-HLTH-SCI-L1B 3 ruin . cvo EpMsmieiogfc Study of Vinyl Chloride Worfcm TABLE I. Dfetrlbtitioa of VIM Stain of Carnot FVC Stady Cohort Members No. Subject to update Alive (as of 12/31/19) Dead (ss of 12/31/19) Identified in oripnal nady (707) Identified is update (129) Unknown (as of 12/31/19) Tool (with updated vtai mm) 6939 1336 723 9200 73-42 16.70 7.88 too Net subject to update AIN* (as of 12/31/1972} Unknown (as of 12/31/1772) Total (not subject to epdate) 935 98.15 18 1.83 973 too 321 The vital status of the cohort as of 31 December 1982 is given in Table I. As of the end of the study period. 6.939 among those with updated vital status were still alive. A total of 1,336 individuals wexe identified as having died during the study period. Among these, death certificates were obtained for all except 97 deaths (6.3 of all deaths). Of the 1,336 deads. 707 were identified in the original study and 829 in the update. A total of 973 individuals were not subject to follow-up in the update, either due to the lack of personal identifiers (the identification codes of 170 were lost) or non-participation of the plants (803). Among those with updated follow-up, the vital status (as of 31 December 1982) of 725 individuals (7.88%) remained uotoown. Of the 973 individuals not subjected to a follow-up in the update. 935 were known to be alive on 31 December 1972, the end day in the original study. As such, personyears up to 31 December 1972 for these individuals were counted in the analysis. Cause-Specific Mortality Anatyals As mentioned earlier, race was not available for all Individuals from the pre vious study. However, there were very few known non-whites (3%) among the employees at the participating plants. For the analysis, die entire cohort was assumed to be white. In the study, a total of 1,536 deaths were identified. Death certificates were obtained for 1,439 (94%) decedents. The 97 deaths without death certificates, but with dates of death, were coded as cause unknown. These deaths were included in the calculation of overall SMRs, but not In any cause-specific SMRs. The observed and expected deaths by cause, SMRs and their 93% confidence limits for the total cohort are presented in Table IT. The 1.536 deaths observed produced an overall SMR of SO.07, which was significantly low (p < 0.01) in comparison with the U.S. national mortality experience. Mortality from cancer of all sites combined was slightly, but not significantly, higher than the U.S. experience (339 observed vs. 341.73 expected. SMR 103.1). The excess came primarily from the following sites: cancer of the liver and biliary tract (hereafter referred to as "liver'') and cancer of the brain and other central nervous system (CNS). For cancer of the liver, 37 deaths were observed, compared OCT PPpri CLEP-HLTH-SC I -L I B _PQS . 000 r Uppar llnm s 94.7 t 116J 5 1*1.3 2 135.1 1 mi 0 114.8 4 107 4 3 8*4.4 7 141.2 6 113.0 7 U5.5 4 398.* 9 177.5 l 1*2.9 1 131.1 .7 217.3 .1 270.* 140.7 .0 246.7 1*1.0 .6 159.4 .4 219.5 .4 201.3 .5 141,7 .4 192.2 96.3 .3 95.2 *92 .4 101.0 a 77J .* 243* 80.* -.4 SA.S U *3.7 .5 86.5 :.9 86.9 1.8 94.8 l.S 116.3 and other 0.05). articular, 15 drathi expected csi found 9 was as Epfdemiofctfc Study of Vinyl Chloride Worten 323 There was a significant deficit in mortality from ail diseases of the circulatory system. The number of observed deaths was 635, compared with 721-28 expected (SMR - 88.0, p < 0.01). The deficit came primarily from arteriosclerotic heart disease (SMR - 81.6, p < 0.01). Although mortality from ail non-malignant respiratory diseases showed a slight nonsignificant deficit, the deficit from pneumonia was statistically significant (15 observed. 31.94 expected, SMR - 47,0). On the other hand, emphysema/chronic obstructive pulmonary disease (COPD) showed a statistically significant excess. The number of observed deaths from empbysema/COPD was 41, compared with 22.83 expected (SMR - 179.6, p < 0.01). A number of statistically significant deficits were observed in the following causes of deaths. For all diseases of the digestive system, 60 were observed, com* pared with 95.58 expected (SMR = 62.8. p < 0.01). The deficit came mainly from cirrhosis of the liver (35 observed, 56.06 expected, SMR s 62.4). Mortality from diseases of the genitourinary system was significantly low (SMR * 44.1, p < 0.05). A statistically significant deficit was also observed in the category of all external causes of death. For this cause of death, 168 deaths were observed, compered with 225.97 expected (SMR " 74.3. p < 0.01). The deficit for external causes of death came mainly from the deficit in all accidents (108 observed, 150.16 expected, SMR - 71.9). 4 Analysis by Length of Expoaura We analyzed SMRs for the entire cohort by length of vinyl chloride exposure by these categories: less then 10 yean. 10-20 yean, and greater than 20 yean (Table m). So obvious increasing trends by length of exposure were identified fa* any came of death, except for cancer of the liver and perhaps cancer of the brain and other CNS. For employees with less than 10 yean of exposure, 6 deaths were due to cancer of the liver, with 3.30 expected (SMR = 182.1, not significant). However, for those with 10 to 20 years of exposure, there were 20 observed deaths due to cancer of the liver, with only 1.62 expected (SMR - 1235.6, p < 0.01). For those employees with more than 20 vean exposure. 11 deaths were due to cancer of the liver, compared with only 0.86 expected (SMR - 1284.9 p < 0.01). For cancer of the brain and other CNS. among those employees with less than 10 yean exposure, 13 deaths were observed and 7.90 expected (SMR 164.7, not significant). Among those with 10 to 20 yean exposure. 4 deaths were due to cancer of the brain and other CNS, compared with 3.31 expected (SMR 120.8, not significant). However, for those employees with 20 or more yean of exposure, there were 6 deaths, compared with only 1.55 expected (SMR * 385.9, p < 0.05). For emphysema/COFD, among those employees with less than 10 yean of exposure there were 25 observed deaths, while 11.97 were expected (SMR " 208.9, p < 0.01). For those with 10 to 20 yean of exposure, there were 11 deaths due to emphysema and COPD compared with 6.59 expected (SMR 166.8, not signifi cant). For the group with more than 20 yean of exposure, the number of emphysema and COPD deaths was S, with 4.27 expected (SMR * 117.1, not significant). Analysis by Latency Long latent periods are usually required for chronic diseases to develop. In many situations, it is more appropriate to examine mortality experience only after a GENC 00538? FROM CLEO-HLTH-SCl-t-IB PAGE.010 nth rfVtaff 20 -r yen Obcnred SMB 194 11.0* 69 11*3 0-- 27 i*o.r 0 W4f I 43.3 1 169.4 11 134.9* a 131.4 19 71.0 19 74.6 0 0 3 104.6 0-- 1 70.2 6 38J.9* 5 104.1 0-- 0 ^P2 110.4 3 214.4 0 2 56.4 0-- 12 S4.3 <4 73J* 37 60.1* 7 4|.2* 1 22.3 9 117.1 4 32.4* 2 27JT 0-- 6 37.9* 2 20.0* 1 24.9 4 93.1 e by these jle IV). For those ;er deaths, h 20 to 30 >90.1, p < J(SMR * Epidemiologic Study of Vinyl Chloride Workers 32S TABLE IV. Observ'd Deaths by Came and SMBs fee All Cohort Members by Latency Knee nut Exposure to Vinyl CMoride____________ Latency tiaca fan arpnsaro______________ < 20 yearn 20-30 yean 30 yean of dead) (7th 1CDA) Ob Ofe Qb- ________________ sermd SMB served SMB aatvad SMB All causes (1-999) AH cancan (140-205) 644 *0.7* 333 103.5 337 90.8 130 93.2 133 113.4 96 113.0 Cancer of twceal cavity and pbwynit (140-141) 6 136-5 4 93.8 2 73.9 Cancar of digestive system (130-139) Cmcer of esophagus (130-130) 31 65.3 2 66.3 32 103.1 2 67.3 36 160.8* 3 143.5 Cancar of stomach (151-130 4 54.7 4 77.9 2 56.1 Cancer of l*gt iatastlns (153-153) 1 72.7 3 30.2* 9 114J Cancer of liver and biliary tract (133--136) 10 383.7* 11 390.1* 16 1218.3* Caneci of pancreas (157-157) 3 70.0 7 107.6 4 84.2 Cancar of tcapuatoty tyuam (160-164) 43 100.0 47 106.1 23 71.6 Cancer of hmg (162-163) 44 KB.B 43 105.0 22 71.8 Cancar of boas (196-196) 0 -- . l 197.4 1 342_3 Cancer Of akin (190-190) 2 49.4 2 92.4 2 173-5 Cancer of promts (177-177) i 27.3 6 113.8 8 127.1 Cancar of bladder (181--19(3 1 36.2 2 67J 2 73.3 Cancer of kidney (180-110) 5 131.8 2 2J 4 193.9 Cancer of Ms and CNS (193-193) 13 163.8 6 138.3 4 210.7 Lymphatic and hematopoietic cancer (200-209) 16 67.4 14 1296 7 97.2 Lymphosarcoma sad mbcuiowcotm (200-200) 7 170.2 3 121.2 I 72.1 Hodgkin's Disease (201-201) 2 31.5 l 90.9 9-- Leukemia and ataifceroia (204-204) 3 71.0 6 148.1 2 70.4 Other lymphatic tsvue caneer (202-203X205-205) 2 63.7 4 134.8 4 176.6 Benign neoplasms (210-239) 2 71.8 2 142.4 0-- Diabetes oeilitus (260-260) 12 113.3 7 94.1 3 54.4 Diseases of blood (290-299) 1 30.7 1 97.1 0 --> Vascular lesions of CNS (330-334) 23 66.1* 20 66.1 27 100.4 Diseases of circulatory system (400-468) 230 63.7* 231 101.7 134 76.T Arteriosclerotic heart discern (430-420) 201 83,9* 200 93.7 87 38.4* Nonmilignam respiratory disease (470-327) 20 42.4* 26 88.2 24 92.0 Poenmonis (490-493) $ 35.6* 3 30.7 3 62.3 Emphysema, including C.O.P.D. (327-327) 11 157.5 17 198.0P 13 179.0 Diseases of digestive system (530-367) 27 34.er IS 49.3* 18 114.6 Qrrboajs of liver ($61-361) 14 4U* 9 48.0* 12 142.1 Diseases of graitouhnsy system (390-639) 1 1.9* 6 U6J 2 30.3 Accidents, poisonings, sod violence (800-991) 123 74.3* 37 14.7 8 47J* AccidanB (800-962) ,. 82 73.2* 18 66.0 8 74.0 Motor vehicle accidents (810-633) 42 73,8* 7 61.9 3 77.9 Suicide (963-963X970-979) 29 79.6 17 143J 0-- Sicnificant at 1% level. ^Significant M 3% level. For cancer of the brain and other CNS. the tread by latency was not at ail obvious. The brain cancer SMRs were 1S3.8, 158.3. and 210.7, respectively, for those employees with less than 20, 20 to 30, and 30 or more yean of latency. Furthermore, none of the three SMRs by latency was statistically significant. The analysis by latency for emphysema/COPD also did oot show any obvious GENC 00539 OCT 3 '91 9:07 FRQn CLEU-HLTH-SCI-L1 a r HUE ,lll( ' the middle 1 groupings: f age (Table 1, 813. and ct depended id by age at ctive SMRs t at tbe 0.05 itgeatftm last group'* edfic SMR xposurc for iver cancer, ficant at the Thus, liver hat of those id 256. with e respective tat the 0.05 vith year of - 18 plants s producing ts for those iduced only nber of tbe only I liver s SMR was There wen aumber of -7 (SMR - respiratory it it was not respiratory ; were only deaths ex- EpMemioiogfc Study of Vinyl Chloride Worktn 327 TABLE V, Observed Deaths by Cam and SMR* tx All Cebert Member* by Age m ft* Erpowit to Vinyl CMorids __________ Age at flm srtpowfs < 39 years 23-34 years 39 * yearn CeoMof dHth (7th 1CPA) "o Ob- Ob- imvca SMR served SMR served SMR All causes (1-999) All cancan (140-309) Cancer of buccal cavity and pharynx (140--14S) Cancer of digestive system (ISO-199) Cancer of oophagu* (150-130) Cancer of uoraach (191-131) Cancer of large imessae (193-193) Cancer of Uver and biliary net (195--156) Cancer of pancreas (lST-t57) Cancer of respiiatecy syiwa (160-164) Cancer of lung (162-163) Cancer of booa (196-196) Cancer of sfcte (190-190) Cancer of picetaK (IT7-177) Cancer of bladder (1S1-1S1) Cancer of kidney (1*0-110) Cancer of brain and CNS (193-193) ___ Lymphatic and bematopoicBC cancer (200-209) Lymphosarcoma and reticokxnreoms (200-300) Hodgkin's Disease (201-201) Leukemia and aieukamia 00*-W) _ Other lymphatic tissue cancw (302-203X303-205) Benign neoplasm* (310-339) Diabetes otelHtns (260-360) Disease* of blood (290--399) Vascular lesioas of CNS (330-334) of circulatory system (400-466) Arteriosclerotic heart discern (430-420) NOemaligsaat rcsprratoty disease (4)0-527) Pneumoaia (490-493) Emphysema, including C.O-7.D. (527--527) Diseaaea of digestive system (530-567) OofcotiS of liver (961-561) Disease* of genitourinary syiscm (990-639) Accidents, poisonings, and viotaoce (600 996) Accidents (100-962) Motor vthidc accidents (610-635) Suicide (963-963X970-979)__________________ __ 149 45 t 12 1 0 0 10 1 15 14 0 2 1 t 0 3 7 3 0 3 1 0 3 0 l 16 6T 7 3 2 14 10 2 57 33 17 16 90,4 947 94.7 131 66J 1134 104.4 -- 9 44- 4 3 --8 1611.0* 16 44.7 6 935 33 91.6 -- 1164 90 2 4 1244 2 142.6 --* 1 6 110.6 9 99.7 13 207.6 3 l 113.0 6 61.0 _ 3 3 63.9 -- 9 0 io.r 23 96.1 196 93.0 149 694 19 113.9 2 106.3 12 61.7 22 91.0 12 63.7 1 73.4* 67 62.3* 47 59J* 21 101.1 17 66.1* 740 93.1 1174 163 96.4 111.6 6 107.9 1374 43 943 1341 2 46.9 94.6 7 60.1 77.1 12 10.7 8134* 11 343.7* 66.0 9 944 1114 47 60.1 1104 47 *4.7 2NJ 0-- 1297 0 -- 49.7 12 115.7 374 3 59.1 169.0 5 119.0 1644 11 239.5* 103.4 15 99.9 9*4 5 149.4 43.6 2 130.3 1134 4 66.6 1534 4 1048 t43J l 49.0 10*4 10 112 2 114,6 604 46 *4.8 76.0* 351 94,1 66.* 272 86.3* 63JP 44 93.0 16.1* 8 4*y 160.0 27 200.7* 35.9* 24 61.4* 49,1* 13 63.0 144* 6 56.5 71.3* 44 8L0 76.4 28 743 746 14 96.4 75.3 13 90.3 *5(gnl6cam at 1% level. 'Significant at 9* level. pected (SMR 219.5, not significant). Far the cohort of PVC employees, 16 deaths were due to brain cancer and other CNS, when only 9.47 were expected. Tbe cor responding SMR was 169.0, p < 0.05. Only one cmpfaysema/COPD death was observed among tbe VCM employees. On the other hand, a total of 29 deaths were due to empbysema/COPD among the PVC employees, compared with 17.75 expected (SMR * 163.4, p < 0.05). OCT .3 '91 9:00 FROM CLEU-HLTH-SCI-LIB PAGE.014 m of Fine me 1900 + ObMPAQd SMR ` 220 74.S* 52 94.4 3 170.0 11 6.9 0 1 49.0 A 93.1 3 41S.9 3 109.7 19 96.5 IB 96.0 0 0 0 0 3 19*. 1 7 256.3* 5 70.5 2 137.4 1 74.3 2 74.7 0 *e. 0_ 3 77.2 0 ____ 7 61.7 67 65.2* 54 62.6* a 66.2 2 42.7 5 116.8 6 32.7* 4 32.7* I 39.5 52 74.4* 33 74.5 IB 76.7 13 (3.3 I I I I 1 1 I 1 I I 1 1 1 1 I 1 1 I 1 I 1 I 1 1 I 1 1 1 I 1 1 I I I I 1 I I I 1 1 of tfaeCMA oss some of i. The ume* i and devtl- EpidamMogic Study of Vinyl Chloride Worker* 329 TABLE vn. Obetrod Deaths by Cease cad SMRa for AH Cohort Mowhsra by Typo cf naat Type of pleat (7th 1CDA) All caosH (1-999) All cancan (140-205) Cancer of buccei cavity tad pharyn* (140-14a) Cancer of digestive system < I50-IS9) Csnccr of esophagus (150-150) Ctoow of stomach (151-iSO Cancer of lane inteetfcw (153-153) Oncer of liver and biliary net (155-156) Cancer of pantsum (157-157) Cancer of tespnasery symn (160-464) Cancer of hag (162-163) Cancer of bone (196-196) Cancer of shin (190-190) Cancer of prosaw (177-177) Cancer cfbMdm(ltl-lSI) Cancer of lddnay (1K>-180> Canev of bnln and CHS (193-493) Lymphatic and hwnatopoiedc cancer (200-206) Lymphewreotaa and resictiioaweQ>M (200-200) Hodgkin's Disease (201-201) Leukemia and sieukaaia (204-204) Other lymphatic otmm emeer (202-203X205-205) Bcaip neoplasms (210-239) Ditbewa mdlitus (260-260) Diseases of blood (290-299) Vascular tesloos of CNS (330-334) Diseases of circulatory systam (400 *6>) Arterfoscierotie heart disease (420-420) Ncnmalignant respiratory disease (470-527) Pneumonia (490-493) Emphysema, indudiuf C.O.tJD. (527-327) Diseases of digeativa sytsem (530-5*7) Cirrhosis of liver (561-5S1) Disoascs of tmxoarray system (590-639) Accident*. poisoning. and viokuec (BOO-99B) Accidents (800-962) Mosor vehicle acodems (B10-S3S) Suicide (963-963X970-979) *Sipiflcaoi sc 1% level. *51gmfkm* as 316 ievaL VCM Observed SMR 76 71.8* 24 113.9 3 434.3 4 77.8 1 206.8 0-- 0 1 325.8 1 91.0 12 135.7 11 150.1 0-- 0 1 144.2 0 1 172.7 2 219J 0-- 0 0-- 0 0-- 0* 0_ 0 3 62.6 36 62.0* 18 51.1* 1 20.4 0 <_ 1 82.6 4 61.9 3 72.6 0 14 77.3 10 85.9 4 68.8 4 95.0 pvc Obserrod SMR 1130 m 6 71 6 7 16 31 11 85 12 I 6 12 5 8 16 31 10 3 11 7 3 14 1 60 463 357 54 11 29 46 26 6 125 80 36 34 J* 104.6 68.5 113.8 97.6 36.4 TLA 693.3 71.3 91.8 93-5 72.8 110.6 98.6 75.6 116.9 169.0 113.6 ` 167.4 74J 104.3 110.8 78.0 77.9 34.2 <3-3 83.6* 78.<T 79.3 44.3* 163.4* 64J* 62.9* 37.4* 75-6* 72.4* 6,6* 882 opmeot of lung cancer, as well as lymphatic and hematopoietic cancer. The update extended the observation period from the end of 1972 to the end of 1982. In w^inn" to the 707 deaths identified in the original study, 829 deaths were identified in d* update, 1973-1982, bringing the total number of deaths in the entire study to 1436. With these newly identified deaths, we now can answer some of the questions that were raised in previous studies. The cohort as a whole continued to experience a favorable general mortality GENC 005395 OCT s; ea F Ron Itcv-rli. I n- 'ij; , u . a ., i number dy, 37 liver r sinulv in dby Beaudicated that perienced a been over* since 1979. nt by plant is demonchanges in s of latency i a marked , could have , the study nalyze liver data by age r risk in our exposed to 9 tunes the de after age risk of liver exposed to wise as the after I960, (founded by id an earlier jfore. these ed on death roma cases 4 angiosar* mrtietpating s of angiorhecoacenexposure at ild tie in the ana would 30*oc tests. cess even if certificates, uciooma. Z r cancer of otnmon bile ircinoma of Epidemiologic Study of Vinyl Chloride Workers 331 the biliary system). Excluding the 15 angiosarcomas, there wen 22 cases of liver and biliary tract cancers with only 5.7 expected, for an SMR of 386.0 (p < 0-02). The above cited liver cancer diagnoses based on death certificates probably included cases of angiosarcoma which were not property diagnosed. In an effort to validate the death certificate diagnosis of the 37 cases of livex/bilsaiy tract cancel, requests for medical records wen made to hospitals and attending physicians. Be cause of the lack of consent from the families, records wen obtained for only 17 cases. In 14 of the 17 cases, actual pathology reports wen available. Among these 14 cases, based on the death certificate diagnosis, there were 8 cases of primary lived biliary tract cancer, 5 cases of angiosarcoma, and 1 case of secondary liver cancer metastatic from another organ. In the same 14 cases, based on pathology records, then were 5 cases of primary ljver/biliary tract cancer. 7 cases of angiosarcoma, and 2 cases of secondary liver cancer. Therefore, in 3/8, or 63% of the subset of 14 cases, the death certificate diagnosis of primary cancer of liv/biliary tract agreed with the pathology records. Among the total 37 cases of liver and biliary tract cancer, 22 were listed 00 death certificates as dying of primary liver/biliary tract cancer, it is likely that cases of angiosarcoma and metastatic cancer of the liver were included among these 22. As indicated by our review of 14 pathology reports, the agreement rate between death certificates and pathology records for the diagnosis of primary cancer of the lived biliary tract was 63%. We can estimate that the correct number of primary lived biliary tract cancers is therefore 22 x 63%, or 13.9. With the expected number of liver cancer deaths of 5.7, the corrected liver cancer SMR is 243 (p < 0-01). Thus, after adjusting for the over-reporting of primary liver/biliary tract cancer, there re mains a significant excess of primary cancer of the liver/biliary tract among this cohort of vinyl chloride wodoea. Another approach was to check the 37 cases of liver/biliary tract cancer against the list of angiosarcomas maintained by Id. According to the IQ list, there were 21 angiosarcomas among these 37 cases, leaving 16 cases of livedbiliary tract cancer. The corresponding SMR for livedbtliary tract cancer would become 281 (p < 0.01). While angiosarcoma of the liver was the main concern as the primary cancer associated with vinyl chloride exposure, there was also an excess of cancer of the biliary tract (including gall bladder) among this cohort. The excess persisted even after adjustments for misdiagnosis based on death certificates were made. According to the Third National Cancer Survey incidence data, 52% of all hepatobiliary cancers are liver, while 48% arise from the biliary tract According to these figures, in our study 2.7 cases (non-age adjusted) would be expected to be from the biliary tract whereas 7 cases were repotted for an observed/expected ratio (in %) of 259 (p < 0.05). This increase in biliary tract cancer (independent of liver cancer) in vinyl chloride workers has not been previously noted or reported. These results are shown in Table vm. , This update confirms the excess in cancer of the brain and CNS. In the entire study, 23 deaths were due to brain cancer, compared with 12.76 expected. The corresponding SMR of 180.2. statistically significant at the 0.05 level, was very similar to the SMR of 203 previously reported for the same cohort by Cooper and the combined relative risk for overall brain cancer of 1.74 estimated by Beaumont and GENC 005397 OCT 9! 9:11 FROn CLEU-HLTH-SC i-u i a * /Wb.. 'Hat. id j o tr mortality anccr tnprl who ware 50 and at a U mortality 1 point. 2 and other to 41 yeanthe largest <ur observa* * number of rthcr psrtic- ide workers f the entire ected 22-83 less than 10 D (SMR t exposed at ijnificant at rred amoag h$ were due iyl chloride reeis*D mortality nvestigator* xposuiea 9 character Lilts. 1980. In addition* Koniosia in jtal. [1975] VC exposed j the present rophysema/ EpManfefogfc Study of Vinyl Chlorid* Wtan 333 COPD excess. In particular, we do not have information on smoking history of the cohon members. Therefore, cigarette smoking could not be assessed is an explana tion for the excess. However, it must be pointed out that there was no excess in lung cancer, which would imply that the cohort did not have higher percentage of smokers than the general population. One of the unresolved epidemiologic issues is the relationship between occu pational exposure to vinyl chloride and respiratory cancer. Previous studies have provided conflicting results on the issue of whether employees exposed to vinyl chloride experienced a higher risk of respiratory cancer. Many of the previous stud ies, which did not find an excess in lung cancer, have been criticized for lack of statistical power. Our study, being one of Che largest studies of vinyl chloride work ers. did not indicate any excess in respiratory cancer. Overall, oar study had 80% power to detect an increased risk as small as 24%. Some of the previous studies have alio shown an increased risk of cancer of the lymphatic and hematopoietic system among vinyl chloride workers, while others did not find such an excess. In our uudy, the corresponding SMR was 102.0. Thus, we did oot detect any excess of these types of cancer. Overall, our study had 80% power to detea an increased risk in lymphatic and hematopoietic cancer as small as 45%. It should be pointed out that there are several limitations in this study, most of which are typical of an historical mortality study of industrial populations. The major limitation of this study is the lack of exposure information for the cohon members. As such, analyses by detailed exposure were not feasible. Another limitation of the study was that a small percentage of the cohort was lost to vital status foilow-up. Among those who participated in the update, the vital status of 7.88% cohort members remained unknown as of December 31. 1982. In addition, among those who did not participate in the update, the vital status of 1.85% cohort workers remain unknown as of December 31. 1972, the end of the observation in the original study. A related limitation was that for 6% of all deaths identified in the study, no death certificate was available. Therefore, it was possible (but not very likely) that the missing information in vital status or cause of death could have modified the results of this study. In the study, the entire cohort was resumed to be white, even though some might be non-white. Among those with race information, only 3% were non-white. However, if the actual number of non-white was larger, some of the results of this study might have been modified. finally, being a mortality study, this investigation not only inherited all the problems associated with death certificates (for example, diagnostic accuracy), but also suffered from the lack of in-depth clinical information. In particular, histologic information on the cell type of the liver cancers and brain cancan would have been useful in interpreting the epidemiologic findings. REFERENCES Beaumont 17, Broalow NB Power conideabea* in tptdemioJopc meGee of vinyl chloride worker*. Am J Sjridetnioi 114:725-734. Bufflci Pa, Wood S. Ether C, Sums L. KSien 07 0979): Mortality expofenre of worken is vtnyt chtonde monomer production plant, t Occnp Mod 21:195-303. Byrea D. Eogboim G. Eitfiuad A. Wemwhohn P (1975): Mortelity aod cancer morbidity in a poop of Swedish VCM and PVC production worken. Eow Kith Pwspect 17:167-170. GENC 00539? 4 Vista Chemical Company 900 Threadneedle Houston, Teias 77224 l Post Offic* Bo* 19029 /Ol- yJA*jL J.R, Drvmwnght, M,0. Martoger-Medical November 8, 1991 Roy T. Gottesman Vinyl Institute Wayne Interchange Plaza n 155 Route 46 West Wayne, N.J. 07470 Dear Roy: Bob Hinderer and I have discussed the Wong et al 1991 study entitled "An Industry-wide Epidemiology Study of Vinyl Chloride Workers, 1942-1982". I first heard about the updated report while attending the Denver meeting of the VCSA. The BNA report had addressed some of Wong's conclusions as they were reported in the American Journal of Industrial Medicine. My comments essentially agree with those already presented to you by Bob Hinderer. Namely, the use of the ICD-7 classification lumps all brain and other CNS cancers together; death certificate data without adequate histopathological diagnosis/review and assumptions established without specific cause/effect, consistency in latency period, etc. Bob's recognition that the authors had apparently failed to have the CMA Panel review the report prior to publication by Wong led to his discussing this with Dr. Shah. It is my understanding that CMA will be addressing this with Dr. Wong. In any event, the study has been published and we must be cognizant of this when new litigation appears. Hopefully, clarification by Wong et al will receive adequate publication as well. GENC 005401 I plan to keep in contact with Bob and will keep you posted as well on any events dealing with this issue. Sincerely, . Drumwright, M.D., FACOM cc: JRD Vi File T. Grumbles R. K. Hinderer, PhD Mgr. Health & Tox. B.F. Goodrich Co. 3925 Embassy Pkwy. Akron, Ohio 44313-1799 GENC 00540 BFGoodrich The BFGoodrich Company 3925 Embassy Parkway Akron. Ohio 44313-1799 October 21, 1991 Dr. Roy T. Gottesman Vinyl Institute Wayne Interchange Plaza II 155 Route 46 West Wayne, NJ 07470 ( Dear Roy: In response to your request, you should note that there are two main problems with the study by Wong et al. 1991, entitled "An Industry-wide Epidemiology Study of Vinyl Chloride Workers, 1942-1982". First, the authors use the 7th rather than the 9th revision of the International Classification of Disease. The effect here is to lump all brain cancers together. Secondly, the authors leave the question of whether there is a relationship between VCM and brain cancer hanging without providing some logical comments and assessments. While they reference Doll's earlier review paper in the introduction, they fail to mention what is important from that paper. Specifically, they fail to note that for brain cancer, and in fact all cancers except liver, there is no consistency across studies which is an important consideration in assessing causal relationships in epidemiology studies. Furthermore, the effect of other potential biasing/confounding factors (eg. other exposures) are not addressed. When I obtained a copy of this publication, I quickly realized that this was the work that was sponsored by the CMA VCM Panel. Unfortunately, the authors failed to send a draft to the Panel for review prior to publication. When I discussed this with Dr. Shah at CMA, he indicated that CMA would notify them that they have violated their contract. The next step will be for the whole Panel to consider what corrective actions are necessary and/or appropriate. Hopefully, the authors will be agreeable to add some clarifying comments in a letter to the editor. 6ENC 005403 1 t Dr. Gottesman October 21, 1991 Page 2 At present, I view this as a CMA project. However, I will keep you posted as things progress. Best regards, THE BFGOODRICH COMPANY Robert K. Hir.derer,Ph.D. Manager, Health and Toxicology Environment, Health and Safety Management Systems 1021-4/jp cc: Dr. Drumwright Dr. Shah - CMA 6ENC 005404